132 Natural History Bulletin. 



up of superimposed lamellcie of structureless horny plates 

 (PI. Ill, fig. 5, a). 



As this species seems a favorable one for demonstrating the 

 origin of the axis cylinder, this subject will now be taken up 

 and the reasons given for the opinion expressed on page 122, 

 to the effect that it is endodermal. 



If the tip of a branch of this coral be embedded in celloidin 

 and transverse sections be made across it, it will be found that 

 the bottom of one of the calicles situated near the tip, not 

 necessarily the terminal calicle, will embrace the extreme tip 

 of the axis cylinder of the branch, or else several calicles will 

 open below into a common chamber which contains the tip of 

 the axis. The former case is illustrated semi-diagrammatically 

 in PI. Ill, fig. 7. This chamber, whether it is the continua- 

 tion of a single calicle or common to several calicles, is lined 

 by a continuation of tlic endothelium of the calicle^ and this 

 endothelial lining of the calicle envelops the growing tip of the 

 axis cylinder and extends downward as an endodermal invest- 

 ment of the axis cylinder, as illustrated in PI. Ill, fig. 7« 



This endodermal investment is doubtless the one referred to 

 by Balfour as having been discovered by Koch.' Koch, how- 

 ever, believed it to be cpiblastic in origin, while all the evi- 

 dence which I have secured from examination of numerous 

 sections, goes to show that it is hypoblastic or endodermal in 

 origin. This investment, moreover, is not furnished by the 

 walls of the water-vascular canals as is the opinion of Colton, 

 there being a considerable layer of mesodermal tissue between 

 the canals and axial investment. 



In fig. 5, PI. Ill, which represents a longitudinal section of 

 the axis cylinder and the structures on one side of it, we see 

 first the horny core (a) with its superimposed lamellae, then 

 the fibrous sheath (e) which seems to be of the same nature 

 as the core, but is not lamellate, then the endodermal invest- 

 ment of the axis (b) which is a continuation of b, fig. 7, then 

 comes a layer of mesoderm, (c, c) with its characteristic spic- 



1 See page 126 of this Bulletin. 



